Text recycling (hereafter TR), sometimes problematically called "self-plagiarism," involves the verbatim reuse of text from one's own existing documents in a newly created text -- such as the duplication of a paragraph or section from a published article in a new article. Although plagiarism is widely eschewed across academia and the publishing industry, the ethics of TR are not agreed upon and are currently being vigorously debated. As part of a federally funded (US) National Science Foundation grant, we have been studying TR patterns using several methodologies, including interviews with editors about TR values and practices (Pemberton, Hall, Moskovitz, & Anson, 2019) and digitally mediated text-analytic processes to determine the extent of TR in academic publications in the biological sciences, engineering, mathematical and physical sciences, and social, behavioral, and economic sciences (Anson, Moskovitz, & Anson, 2019). In this article, we first describe and illustrate TR in the context of academic writing. We then explain and document several themes that emerged from interviews with publishers of peer-reviewed academic journals. These themes demonstrate the vexed and unsettled nature of TR as a discursive phenomenon in academic writing and publishing. In doing so, we focus on the complex relationships between personal (role-based) and social (norm-based) aspects of scientific publication, complicating conventional models of the writing process that have inadequately accounted for authorial decisions about accuracy, efficiency, self-representation, adherence to existing or imagined rules and norms, perceptions of ownership and copyright, and fears of impropriety.